What Happens to Citizen Complaints Against Cops in New York?

An analysis of complaint data from the Citizen Complaint Review Board
By Isabelle Chun, Meital Hoffman, Dain Kim and Nicolas Rothbacher

Introduction

On May 25th, Derrick Chauvin, a white Minneapolis Police Officer, murdered George Floyd, a Black man suspected of using a counterfeit bill, by kneeling on his neck for over eight minutes.¹ The death of George Floyd sparked an international outrage. Minneapolis Police Department records show that civilians had filed 18 complaints about Chauvin over the course of his career.² Activists question the effectiveness of existing accountability mechanisms for police misconduct.

Protesters gathered around the world chanting: George Floyd, Breonna Taylor, Ahmaud Arbery, Tony McDade. All names of Black Americans murdered by police officers in 2020. These events, combined with the blaring racial disparities in COVID patients,³ brought institutional racism to the forefront of the American conscience. Despite decades of failed efforts at police reforms, this movement, characterized by continuous protests and calls to defund the police, has actually precipitated promising policy changes aimed at ensuring transparency and accountability for the police as well as reducing police violence.

Here is where we turn our attention to New York, the focus area for our research. In the video of George Floyd’s death he tells the officers: “I can’t breathe”. These words are unfortunately familiar to many in New York. In 2014, the New York Police Department (NYPD) used a chokehold against Eric Garner, a Black man suspected of selling loose cigarettes, resulting in his death. Eric Garner is heard on video repeating the words, “I can’t breathe”. After the first week of protests following George Floyd’s death, the New York State Assembly passed a statewide ban on chokeholds, titled the “Eric Garner Anti-Chokehold Act”.⁴ Some viewed this measure as purely symbolic, since New York City had a ban on chokeholds in effect when Eric Garner was murdered. Also on June 6th, the New York State Assembly repealed Section 50-A of the Civil Rights Law. The law had previously prevented any officer personnel records from being disclosed to the public, and was one of the most restrictive laws for public access of police records in the United States.⁵ With the repeal of 50-A, New York has promised to release historical and new data about police personnel and conduct records. The potential new data are discussed later.

While updates to officer training, the incorporation of body cameras and other initiatives are put in place to address police conduct and promote accountability, studies have shown that networks of misconduct in police forces are hard to break.⁶ The norm against reporting fellow police officers for misconduct has been named the “blue code of silence”.⁷ Within the NYPD, former officer Adrian Schoolcraft and the retaliation he faced from other officers for his reporting of internal misconduct reflects the influence of the blue code of silence and challenge in enacting lasting change.

News headlines typically focus on the worst cases of police brutality, and many instances of excessive force, biased treatment, and invasive stop-and-frisks go unpunished and unrecognized by police leadership. Filing complaints is one way for civilians to report incidents of police officers overstepping their authority. For cities, complaints can reveal sources of tension in the relationships between police officers and local communities. In this report, we analyze a particularly long-standing and large dataset of anonymized citizen complaints from New York City filed through the Civilian Complaint Review Board, an independent investigative body staffed by civilians.

Steps for filing a complaint:

1. File a complaint with the CCRB. Complaints consist of one or multiple allegations against one or multiple officers.
2. Within 48 hours of a complaint submission, the CCRB assigns an investigator who may offer the complainant the opportunity for a face-to-face mediation with the officer(s) involved.⁸
3. For the investigation path, the investigator conducts an in-person interview of the complainant, contacts witnesses, visits the location of the incident, and can use the CCRB’s subpoena power and access to NYPD records to obtain relevant evidence and identify the officer.⁹
   If civilians involved in the complaint refuse to cooperate with the CCRB, the complaint is labeled “truncated” and the investigation concludes.
4. Investigators interview the officer(s) named in allegations following the complainant in-person interview.
5. For fully investigated cases, three of the CCRB’s fifteen board members review all available evidence then decide whether this complaint is substantiated.¹⁰ More information about what the different outcomes mean can be found here. If a complaint is substantiated, the reviewing board members make discipline recommendations to the Police Commissioner.¹¹
   There are three levels of discipline that the CCRB can recommend: instructions or formalized training, command discipline (loss of vacation days), and charges and specifications (loss of vacation days, suspension, or termination). All cases where the CCRB recommends changes and specifications go to an NYPD trial.¹²
6. After this point, it is up to the Police Commissioner’s discretion to follow the recommendations or not.

The CCRB receives thousands of complaints from New Yorkers each year and its team of nearly 200 full-time investigators process “FADO”-type complaints. FADO stands for force, abuse of authority, discourtesy, and offensive language; a complaint concerned with at least one of these would fall under the CCRB’s jurisdiction.

In addition, the CCRB publishes monthly and semi-annual reports, and issue-focused reports in which they study a specific type of complaint then make policy recommendations. As a complaint review board, the CCRB operates independently of the NYPD and lacks the authority to enforce their recommendations both for individual officers’ punishment and for policies and practices affecting the entire police force.

Data Analysis

The three datasets we use include complaints received, complaints closed, and allegations closed. On average, the CCRB receives around 5,700 complaints every year. Figure 1 shows the number of complaints received annually over the course of the entire dataset.

Figure 1: Complaints received by year

Each complaint can contain one or more allegations. For instance, a complaint filed in 2017 made three allegations: use of race-related offensive language, refusal to provide name and shield number, and threat of arrest (Figure 2). Each of these allegations falls under one of four FADO categories. In this complaint, the first allegation was categorized as offensive language, and the other two as abuse of authority. The dataset also includes additional information such as the duration of the case and the borough of incident. On average, there are 2.95 allegations in each complaint.

Figure 2: A few complaints from the data set

Complaint Outcomes

Once a complaint is received, the CCRB decides whether to pursue an investigation, truncate the case, or take a different approach altogether (such as attempting mediation) based on the charges and the evidence present. Figure 3 shows a flowchart summarizing how the allegations are categorized and handled. Roughly half of the allegations are fully investigated and the other half truncated. On rare occasions, mediation attempts are made.

Figure 3: Sankey diagram of the complaint process

Figure 4 shows the number of investigated complaints that result in each outcome. Only a small fraction of the investigated complaints are substantiated: of the 38,102 complaints investigated from 2000 to 2017, only 4,671 (12%) have resulted in substantiation and some form of discipline. The most common outcome was unsubstantiated with 17,206 cases (45%), followed by exonerated and unfounded (both 18%).

Figure 4: Outcomes of investigated complaints

On the other hand, complaints are truncated if the CCRB cannot get the necessary cooperation from the civilians involved while conducting the investigation. As shown in Figure 5, the vast majority of complaints are truncated because the complainant is “uncooperative” or is otherwise unavailable.

Figure 5: Outcomes of truncated complaints.

Why do so many complainants suddenly become uncooperative? One possible explanation lies in the CCRB’s complaint filing process. After filing a complaint, complainants are required to conduct an in-person interview regarding the incident. We hypothesize that this in-person interview requirement is a significant barrier to investigating complaints. Complainants may decline the interview for various reasons, such as fear of retribution or a simple scheduling issue. In order to mitigate these external factors from hindering the investigation, we believe that the CCRB should relax the in-person interview requirement. We recommend this interview be conducted over the phone, or if adequate information is provided in the initial complaint filing, that the interview only be conducted when it becomes necessary. Other strategies to assist people who may otherwise be unavailable for an interview is to extend CCRB hours into the evening or weekend occasionally. Alternatively, the CCRB could conduct better outreach to specific communities, such as targeting the areas that have the most complaints truncated or partnering with community organizing groups to host pop-up interview opportunities.

What Impacts Substantiation?

Substantiated complaints are those which have been investigated and are found to be true. Substantiated complaints may, but do not always, result in discipline of an accused officer. Below are charts of the substantiated complaints split by FADO type and allegation.

Figure 6: Substantiation rate and allegation counts for types of use of force allegations

In figures 6 through 11, we see that few of the allegations are substantiated, or found to be both true and against department policy. In force allegations, “Physical Force” is the only subcategory that has a significant number of substantiated allegations. Force allegations across the entire dataset are substantiated at a rate of 1.45%.

Figure 7: Substantiation rate and allegation counts for types of abuse of authority allegations

Abuse of Authority allegations are the most frequently substantiated with an average substantiation rate of 5.66%. Frisk has the highest rate among the subcategories, where 12.55% of total frisk allegations are substantiated.

Figure 8: Substantiation rate and allegation counts for types of discourtesy allegations

Within the Discourtesy allegations, Word has the highest number of allegations and the most substantiated. The total rate of substantiation is 2.66% for Discourtesy allegations.

Figure 9: Substantiation rate and allegation counts for types of offensive language allegations.

Under Offensive Language, the smallest section of total allegations, 1.92% are substantiated. Only 1.41% of complaints stating officers used offensive racial language were substantiated.

Figure 10. Total and Substantiated Allegations by Year

Figure 10 looks at the total allegations and the total substantiated allegations per year. Overall, the raw number of substantiated allegations remains relatively constant regardless of the total number of allegations. While there were more than twice as many allegations in 2009 than 2001 (25053 in 2009, 9871 in 2001), the substantiated allegations only went up by 38% (1139 in 2009, 824 in 2001).

Figure 11: Substantiation rate over time

Figure 11 explores the rate of substantiation according to year. Although total allegations grew and then declined, about the same number of allegations were substantiated from 2004 to 2008. This is reflected in the shape of the graph in Figure 10, which is essentially the inverse of the total allegations. It is surprising that the rate of substantiated complaints is inversely correlated with the total number of complaints received, as seen in Figure 12. Figure 12 also shows a negative correlation between total allegations and the fraction of complaints investigated and substantiated. There are several potential explanations for these relationships. For example, the CCRB and NYPD may be resource constrained, making it difficult to investigate complaints as more come in. Another possible reason is that there were difficulties in investigating stop-and-frisk, resulting in the lower rates of investigation and substantiation during the peak stop-and-frisk years.

Figure 12: Negative correlation between allegation count and investigation rate (a), the same for substantiation rate (b).

Stop and Frisk

Stop-and-frisk was defined in the United States in the 1968 Supreme Court case Terry v. Ohio. The case set a precedent for the Fourth Amendment interpretation of what qualifies as a search, and when a search is acceptable. Although the lawfulness of discretionary stops on the street was codified as constitutional by Terry v. Ohio, the practice was relatively informal, at least in New York City, until the 1990s. Some former police officers claim that the introduction of the record keeping software CompStat under Mayor Rudolph Giuliani led to the first widespread use of the “stop, question, and frisk” approach to law enforcement in the NYPD.¹³ However, CompStat did not drive stop-and-frisk to its height. When Mayor Michael Bloomberg came into office in 2002, he made it an explicit goal to drive up the number of stop-and-frisks, from the current level of 97,000 per year.¹⁴ At the height of the expansion, in 2011, there were more than 685,000 stops, 88% of which did not result in any fine or arrest.¹⁵ Public outcry, multiple lawsuits and Mayor DeBlasio’s vocal opposition during his mayoral campaign and first months in office forced the NYPD to reduce stops. By 2017, there were just 11,000 stops, a 99% decrease from the height of the practice, and 70% of those resulted in a fine or arrest.¹⁶ The yearly count of discretionary stops is shown in Figure 13.

Figure 13: Stop-and-frisks by the NYPD by year

In the complaint data, complaints are labeled by whether the complaint was related to a stop-and-frisk incident. Figure 14 visualizes the total complaints closed by year, colored by whether or not they are tagged as a stop-and-frisk complaint. We can see in this diagram that stop-and-frisk complaints make up most of the year-over-year changes in complaints during the height of the program.

Figure 14: Complaints colored by whether they are related to stop-and-frisk

Does the increase in stop-and-frisk related complaints totally account for the changes? In Figure 15b, stop-and-frisk related complaints are excluded. Here we can see that while non-stop-and-frisk complaints largely followed a random pattern, stop-and-frisk complaints consistently increased between 2002 and 2007 and consistently decreased from 2010 to the present with the exception of 2013 which saw a major increase in both types. Stop-and-frisk related complaints peaked between 2006 and 2010 which, comparing this with Figure 13, aligns with years when the NYPD were making the most stops. This indicates that stop-and-frisk significantly increased the number of complaints over the years it was heavily used.

Figure 15: Yearly stop-and-frisk complaints (a) and non-stop-and-frisk complaints (b).

It is worth noting that despite the sharp cut-off of stops beginning in 2013, there was a considerable increase in complaints that year. This could be due to a backlog from 2012 caused by Hurricane Sandy which hit New York City in late 2012 causing significant disruption to daily life. It also might be explained by the ruling Floyd v. the City of New York in 2013 in which a federal court ruled the practice unconstitutional. Residents in New York may have become aware that such stops infringed on their rights. Thus, we hypothesize that although fewer stops occurred, a larger percentage of people who were stopped may have filed a complaint, since they knew that the practice was not consistent with NYPD policy. This also might explain why complaints about stop-and-frisk did not drop off nearly as quickly as stop-and-frisk incidents.

Conclusion and Limitations

Our study relied on limited data from a particular organization to investigate their practices and while we found some illuminating results, there is much work needed to further understand these conclusions. First, we found that substantiation is very infrequent and that about half of complaints are never investigated. While the most common justification for not investigating was “Complainant Uncooperative/Unavailable,” more research and data would be required to understand why that outcome happens often. We also need more data to understand why the majority of investigations end in inconclusive “unsubstantiated” decisions. Is there something about the process that makes evidence of what happened difficult to find? Second, we found a significant negative correlation between the number of allegations the CCRB reviews and the rate of substantiation. Additional data and research will be necessary to understand why this occurred. Finally, we found that the Bloomberg Era campaign of stop-and-frisk significantly increased the number of complaints made to the CCRB and that stop-and-frisk incidents were the driver of this increase. New data and powers are already being made available this year with new laws and policies intended to strengthen accountability for the police, we hope that with these mechanisms more truths can be uncovered and reports will produce real change, not get put into filing cabinets.

In November 2019, New York voters approved Ballot Question 2 concerning expansion of the CCRB. The initiative granted the CCRB “the authority to investigate and prosecute when a member of the NYPD has made a false claim before CCRB staff about the incident under investigation during a CCRB investigation”, added 2 board members, and guaranteed the CCRB staff headcount to “0.65% of the number of uniformed officers in the NYPD” (0.59% in 2019).¹⁷ These changes are a concrete step in supporting the CCRB’s authority and expanding the number of groups who hold police officers accountable.

Although the complaint data that the CCRB publishes online spans many years, its level of granularity makes it impossible to test hypotheses relating to locations of complaint incidents, demographics of complainants and accused officers, and changes in the volume of complaints after high profile events and NYPD policy changes. Currently, only the borough and year that a complaint was closed are available for location and time data. When a civilian files a complaint, listing their race(s) is optional and confidential. With more information, we could better understand which neighborhoods and ethnic groups report the most complaints, how complaint outcomes differ, and potentially find other disparities in substantiation rates.

On June 17, 2020, New York State passed multiple bills addressing long-standing issues in policing. Among these bills was the repeal of Section 50-A of the civil rights law in New York and the upcoming release of NYPD officers’ discipline records.¹⁸ The protected status of these records not only enables problematic officers to remain employed, but also does not inhibit promotions or gun-carrying privileges. Mayor Bill de Blasio announced that data on officers’ disciplinary history would be available in mid-July, and the CCRB has released new data that include disciplinary records of NYPD officers and identify officers in complaints.¹⁹ ²⁰ Through this new data, we hope the public and the press will be able to further analyze misconduct history and patterns but were unable to do so in this analysis.

Footnotes:

1. Hill, Evan, Ainara Tiefenthaler, Christiaan Triebert, Drew Jordan, Haley Willis, and Robin Stein. 2020. “How George Floyd Was Killed In Police Custody”.The New York Times https://www.nytimes.com/2020/05/31/us/george-floyd-investigation.html.
2. Dakin Andone, Hollie Silverman, and Melissa Alonso, “The Minneapolis Police Officer Who Knelt on George Floyd’s Neck Had 18 Previous Complaints against Him, Police Department Says,” CNN (Cable News Network, May 29, 2020), https://www.cnn.com/2020/05/28/us/minneapolis-officer-complaints-george-floyd/index.html.
3. Coronavirus Disease 2019 (COVID-19)”. 2020. Centers For Disease Control And Prevention. https://www.cdc.gov/coronavirus/2019-ncov/need-extra-precautions/racial-ethnic-minorities.html.
4. “Assembly Passes Eric Garner Anti-Chokehold Act,” Assembly Passes Eric Garner Anti-Chokehold Act, June 8, 2020, https://nyassembly.gov/Press/files/20200608a.php.
5. “New York State Assembly: Bill Search and Legislative Information,” New York State Assembly | Bill Search and Legislative Information, https://nyassembly.gov/leg/?default_fld=&bn=9332&term=2015&Summary=Y&Memo=Y.
6. Wood, G., Roithmayr, D., & Papachristos, A. V. “The Network Structure of Police Misconduct”. Socius (2019).
7. Chin, G. Wells, S.,“The “Blue Wall of Silence” as Evidence of Bias and Motive to Lie: a New Approach to Police Perjury, University of Pittsburgh Law Review, 59(233), 1998.
8. “Mediation,” Mediation — CCRB, https://www1.nyc.gov/site/ccrb/mediation/mediation.page.
9. “Investigations,” Investigations — CCRB, https://www1.nyc.gov/site/ccrb/investigations/investigations.page.
10. “Case Outcomes,” Case Outcomes — CCRB, https://www1.nyc.gov/site/ccrb/investigations/case-outcomes.page.
11. CCRB Board Meeting Minutes, June 10, 2020 pg 46. https://www1.nyc.gov/assets/ccrb/downloads/pdf/about_pdf/board/2020-meeting-minutes/20200610_boardmtg_minutes.pdf
12. “Police Discipline,” Police Discipline — CCRB, https://www1.nyc.gov/site/ccrb/prosecution/police-discipline.page.
13. Naspretto, E. “The Real History of Stop-and-Frisk”, New York Daily Post, (2012).
14. Southall, A. “Why ‘Stop-and-Frisk’ Inflamed Black and Hispanic Neighborhoods”, New York Times, (2019).
15. “Stop-and-Frisk Data”, NYCLU, accessed: 11/25/2019
16. Ibid.
17. “What’s True on Question 2?:”, City of New York https://www1.nyc.gov/site/ccrb/about/outreach/charter2019.page.
18. “NY State Assembly Bill A10611,” NY State Senate, June 11, 2020,https://www.nysenate.gov/legislation/bills/2019/a10611.
19. “Mayor De Blasio Announces New Reforms to NYPD’s Disciplinary System,” City of New York, June 17, 2020, https://www1.nyc.gov/office-of-the-mayor/news/445-20/mayor-de-blasio-new-reforms-nypd-s-disciplinary-system.
20. Bronck, Jonas. 2020. “CCRB Chair Fred Davie on 50-a Repeal.” The Bronx Daily. June 12, 2020. https://bronx.com/ccrb-chair-fred-davie-on-50-a-repeal/.